READ ME

####################################################################################
#
# Replication materials for: 
#     Ingram, Matthew C. 2016. "Mandates, Geography, and Networks: 
#     Diffusion of Criminal Procedure Reform in Mexico"
#     Latin American Politics and Society 58: 121-145.
#     DOI: http://dx.doi.org/10.1111/j.1548-2456.2016.00301.x
# Author: Matthew C. Ingram
# Affiliation: University at Albany, SUNY
# contact: mingram@albany.edu
# 
####################################################################################

Content of all replication files is from May 2015 or earlier. 

Original manuscript was submitted to journal in September 2014, editors invited revision in February 2015, revised manuscript was submitted to journal in May 2015, editors issued a conditional acceptance in July 2015, and minor revisions were made and final version of manuscript (subsequently accepted) was submitted in August 2015. Thus, no major substantive changes to analysis were made after May 2015. The replication files have some more recent changes, but these are just cosmetic (e.g., comments) to make them more accessible and readable.

There are three types of data files: 
(1) a rectangular dataset in Stata format (.dta), and 
(2) spatial data in the form of shapefiles (.shp), including a simple file with state boundaries in Mexico and another files with data from this study.
(3) spatial weights data in the form of .gal files created in GeoDa; GeoDa is free software available for download (https://spatial.uchicago.edu/software)

There are two types of command files: 
(1) .do files for use with Stata, and 
(2) .R scripts for use with R. 
The main statistical analysis (reported in Table 1 and Table 2) and the graph in Figure 2 were done in Stata. 
Several stages of data management, creation of spatial and network lags, visualizations (Figures 1, 3, and 4), and the generation of the random graph (see article, pages 135-136) were done in R. 
See article and command files for details regarding version information and packages.

For readers wishing to replicate the main results reported in Table 2, the Stata .do file and the main Stata data set (.dta) are all you 
need.
For readers wishing to replicate the figures, the Stata .do file replicates Figure 2, and the main .R script generates all the other figures.

For readers wishing to reproduce all robustness checks discussed in articles, the .do file provides code for alternative model 
specifications and other robustness checks, and the second .R script shows how random graphs and associated data were generated.
